22 thoughts on “Eevblog #10 part 1 of 2 – how a rubber band cost millions of dollars”
be specific.. and not let it be so relative and flamed with confirmation bias.. "we purchase some cheap chinese DMM's"??.. c'mon if these DMMs really was such a problem then you would also remember the models. some Chinese DMMs are cr4p versus the price they are asking.. just as some western DMMs are rubbish versus the price they are asking and vice versa.
That story about company that produce high cost and performnace product and procedure of quality chack involve only one (cheap) multimeter, is crazy. I don't belive it.
I know this is 10 years old, but a similar story happened in a New Mexico nuclear waste storage facility. This is a low level waste repository for like medical and lab waste. It is in a salt dome so once the mine shafts are sealed, the salt over geologic time will encapsulate the waste in salt for millions of years. There was a waste shipment from Los Alamos that consisted of liquid waste absorbed into cat litter. This is the normal process. However, some purchasing manager decided to "go green" and instead of using clay cat litter, as the process spec called for, they used shredded and pelletized newsprint litter. Once the waste was packed into the mineshaft, the newsprint dried out enough to allow the waste to start reacting chemically with other waste and the newsprint (and waste heat from decay) resulting in a smouldering fire in the waste barrels, not just once, but at least twice, before the root cause was identified. Purchasing agents can indeed be quite destructive.
ahHA! Now I see. You werent the super organized (in terms of the workbench) man that we all now know you to be! Thank the Gods! I can now rest easy about how much of a mess my workbench is when I am done with whuteva
Jeez, Even some of the crusty companies/ inventors I have worked for would have done some basic material compatibility checking. Oil for instance will do very bad things to natural rubber. But I have seen weird ass stuff happen with solvents and plastics and "rubber". So no surprise that they had problems. Thing is that even a name company might change the formula they make rubber bands from so using them without a spec. sheet of exactly what the "rubber" actually is was the first mistake. I have used O Rings in similar situations and those you can get in a variety of materials. Cost more up front but…
I'd hate to sound cliche with how often it's mentioned but the Hyatt Regency disaster was a really tragic demonstration of how a seemingly insignificant change can have drastic effects in the end.
This sounds familiar. Today in fact, the company I work for just got a "qualifying run" of ten PCB boards from a new manufacturer. The boards are actually higher quality then the previous vendor could provide, however, the damn things won't pass testing on our bed of nails test. They work functionally, however our chief test engineer is refusing to use them because the bed of nails is failing them for faulty (closed) spark gaps on the board that we cannot recreate. Our best guess currently is an incorrectly laid out ground plain, but we've yet to confirm this. Moral of this story is, when you change something, order small lots and test the ever loving shit out of them. If these boards had passed test, they'd be assembled and sent to specific customers labeled specifically as a field test unit until we can determine if they are operating in a satisfactory way. A similar change was made to an older product previously, and the PCBs passed tests both just post building and in their final assembly, but it turned out that poor materials killed the performance when the PCBs got hot, and in this industry, faulty readings could cost governments millions of dollars and potential lives lost. Had we not sent these boards out as a qualifying run, it would have been a disaster, but thankfully the issue was noted with only ten test units in customer hands and quickly fixed.
At the end, you end up gambling a lot when you buy cheaper products, but you can get lucky. I had a brandless multimeter for about 10 years and it works like a charm.
be specific.. and not let it be so relative and flamed with confirmation bias.. "we purchase some cheap chinese DMM's"??.. c'mon if these DMMs really was such a problem then you would also remember the models.
some Chinese DMMs are cr4p versus the price they are asking.. just as some western DMMs are rubbish versus the price they are asking and vice versa.
My Ingles a Tecnical Linguage !!!
That story about company that produce high cost and performnace product and procedure of quality chack involve only one (cheap) multimeter, is crazy. I don't belive it.
Damn 2009 watching in 2020.
Who is watching in 7/2020
I know this is 10 years old, but a similar story happened in a New Mexico nuclear waste storage facility. This is a low level waste repository for like medical and lab waste. It is in a salt dome so once the mine shafts are sealed, the salt over geologic time will encapsulate the waste in salt for millions of years. There was a waste shipment from Los Alamos that consisted of liquid waste absorbed into cat litter. This is the normal process. However, some purchasing manager decided to "go green" and instead of using clay cat litter, as the process spec called for, they used shredded and pelletized newsprint litter. Once the waste was packed into the mineshaft, the newsprint dried out enough to allow the waste to start reacting chemically with other waste and the newsprint (and waste heat from decay) resulting in a smouldering fire in the waste barrels, not just once, but at least twice, before the root cause was identified. Purchasing agents can indeed be quite destructive.
"Original 87"?
I've got a Fluke 87…without the "V"… but it still has most of the features of the "V"!
Wow, u look so young here
btw, 2019 and STILL an awesome video!
ahHA! Now I see. You werent the super organized (in terms of the workbench) man that we all now know you to be! Thank the Gods! I can now rest easy about how much of a mess my workbench is when I am done with whuteva
Jeez, Even some of the crusty companies/ inventors I have worked for would have done some basic material compatibility checking. Oil for instance will do very bad things to natural rubber. But I have seen weird ass stuff happen with solvents and plastics and "rubber". So no surprise that they had problems. Thing is that even a name company might change the formula they make rubber bands from so using them without a spec. sheet of exactly what the "rubber" actually is was the first mistake. I have used O Rings in similar situations and those you can get in a variety of materials. Cost more up front but…
ahh, humble beginnings.
Why were you using a cheap multimeter when you were making a product of 30 000 dollars
If you build thousands of "x" with a rubber band then you deserve bad reputation, unless x==rubberband 😀
Speaking of rubber bands, I heard of one Thiokol installed in a Space Shuttle SRB once…
I'd hate to sound cliche with how often it's mentioned but the Hyatt Regency disaster was a really tragic demonstration of how a seemingly insignificant change can have drastic effects in the end.
He complains about cheap multimeters, yet his camera is a potato.
Saving money at the wrong spot….. Testing 30k equipment with a 30$ multimeter is simply dumb!!
This sounds familiar. Today in fact, the company I work for just got a "qualifying run" of ten PCB boards from a new manufacturer. The boards are actually higher quality then the previous vendor could provide, however, the damn things won't pass testing on our bed of nails test. They work functionally, however our chief test engineer is refusing to use them because the bed of nails is failing them for faulty (closed) spark gaps on the board that we cannot recreate. Our best guess currently is an incorrectly laid out ground plain, but we've yet to confirm this. Moral of this story is, when you change something, order small lots and test the ever loving shit out of them. If these boards had passed test, they'd be assembled and sent to specific customers labeled specifically as a field test unit until we can determine if they are operating in a satisfactory way. A similar change was made to an older product previously, and the PCBs passed tests both just post building and in their final assembly, but it turned out that poor materials killed the performance when the PCBs got hot, and in this industry, faulty readings could cost governments millions of dollars and potential lives lost. Had we not sent these boards out as a qualifying run, it would have been a disaster, but thankfully the issue was noted with only ten test units in customer hands and quickly fixed.
At the end, you end up gambling a lot when you buy cheaper products, but you can get lucky. I had a brandless multimeter for about 10 years and it works like a charm.
Wow episode number ten. You've gone so far, Dave. Impressive progress.